Orderly and specific connections among neurons of multiple subtypes underlie the function of neural circuits. Our own work has used retina as a model to elucidate molecules and mechanisms that underlie specificity, focusing on the matching of pre- and postsynaptic partners in particular synaptic laminae. There is another sort of order, however, that has received less attention: the arrangement of cells and their neurites in the orthogonal (x-y) plane. Processes that contribute to this arrangement include mosaic spacing of neuronal somata, tiling of dendrites, and self-avoidance of processes within a single arbor. These processes are believed to ensure uniform coverage of the visual field, precise connectivity, and appropriate receptive field size. Their molecular bases remain unknown in vertebrates. Recently, we began analyzing two sets of cell surface proteins that we suspected to be involved in laminar specificity: MEGF10 and 11, and a cluster of 22 related gamma protocadherins (Pcdhgs). Unexpectedly, preliminary results suggest that both are involved in regulating the arrangement of specific neurons and their dendrites in the x-y plane. Moreover, both affect the same cell type, starburst amacrine cells (SACs), but in different ways: MEGF10/11 regulate the mosaic arrangement of SAC somata whereas Pcdhgs are required for self-avoidance of their dendrites. We will now use these results as starting points to obtain insights into the mechanisms that underlie these common but little-studied aspects of circuit assembly. First, we will use gain- and loss-of function methods in vivo to characterize the role of MEGF10/11 in mosaic formation. We will ask whether MEGF10/11 is effective only during development, whether its effects endure, whether it can disrupt mosaics after they form, and whether the two homologues have distinct effects. Second, we will ask whether defects in self-avoidance observed in conditional Pcdhg mutant mice are cell-autonomous and whether they reflect problems in dendrite formation or refinement. We will then use genetic methods to reduce the repertoire of Pcdhg isoforms that retinal cells express. We can thereby test the role of isoform diversity in the process and learn whether different isoforms play different roles. Third, we will ask whether retinal subtypes other than SACs use MEGF10/11 or Pcdhgs to pattern their somata or arbors. Finally, we will combine studies in vitro and in vivo to initiate analyses of the signaling mechanisms by which MEGF10/11 and Pcdhgs function. We will ask whether MEGF10/11 act as receptors, as ligands, or as both ligand and receptor (that is, homophilically). For Pcdhgs, we will ask how the initially adhesive interaction that Pcdhgs appear to promote is translated into the repellent one required for self-avoidance. Together, these results will further our understanding of two poorly understood gene families and of a poorly understood set of processes important for patterning neural circuits.